Morphological profiling of cells using multiplexed dyes, such as cell painting, has emerged as a cost-effective and data-rich modality that when combined with AI for analysis has shown powerful applications in compound profiling and -screening. In my presentation I will present the ongoing work in the Pharmaceutical Bioinformatics (https://pharmb.io/) group at Uppsala University in Sweden, where we have established an automated lab for cell profiling and pursue academic drug discovery and compound profiling. I will describe how we set up our automated lab and associated computational infrastructure with data pipelines and tools, and present some of our recent methods developments, including efforts in AI-modeling of label-free imaging, combining cell morphology with other modalities, and our ongoing work to expand into 3D spheroids. I will also show some highlights from recent applications including MoA predictions and drug screening.
